Rapid Hydrothermal Deoxygenation of Oleic Acid over Activated

Apr 8, 2015 - Department of Civil and Environmental Engineering, Old Dominion University, 5115 Hampton Boulevard, Norfolk, Virginia 23529,...
2 downloads 0 Views 1MB Size
Article pubs.acs.org/EF

Rapid Hydrothermal Deoxygenation of Oleic Acid over Activated Carbon in a Continuous Flow Process Sergiy Popov and Sandeep Kumar* Department of Civil and Environmental Engineering, Old Dominion University, 5115 Hampton Boulevard, Norfolk, Virginia 23529, United States S Supporting Information *

ABSTRACT: In this study, a novel approach to converting fatty acids into n-alkanes was investigated. Fuel range hydrocarbons were obtained in a continuous flow process from oleic acid using near- and supercritical water as the reaction medium, granulated activated carbon as a catalyst, and 1% v/v formic acid as an in situ source of hydrogen. Experiments were conducted in a packed tubular reactor with the weight hourly space velocity of 4 h−1 at temperatures from 350 to 400 °C and pressure 3500 psi (24.1 MPa). The oil to water to formic acid ratio was 1:5:0.05 by volume. The main reaction pathways were hydrogenation of oleic acid and decarboxylation/decarbonylation of the resulting stearic acid to form heptadecane. The yield of heptadecane of above 70% with a selectivity 80% was observed between 370 and 380 °C. The results of the study show that efficient hydrothermal deoxygenation of fatty acids can be achieved with activated carbon as a catalyst and formic acid as an in situ source of hydrogen within minutes. Kinetics study showed that the rates of oleic acid conversion displayed Arrhenius behavior with an activation energy of 120 kJ/mol.

1. INTRODUCTION Biomass with high lipid content (e.g., jatropha seeds, microalgae) is an attractive feedstock for renewable fuel production. There is a considerable interest in research and development of methods for converting triglycerides and fatty acids into hydrocarbons that are fully fungible with petroleum-derived transportation fuels.1 Significant efforts were made to adapt petroleum hydrotreating technology to producing hydrocarbons from plant oils.2 These efforts included providing respective reaction conditions (250−400 °C, 10−20 MPa), using conventional hydrotreating catalysts (e.g., CoMo/Al2O3, NiMo/Al2O3) or noble metal catalysts (e.g., Pt/C, Pd/C), and supplying H2 with high feed rates (100−700 L/L) for removal of oxygen from oils in the form of H2O. The high consumption of gaseous H2 makes the hydrodeoxygenation process costly. In order to reduce the production costs, consumption of H2 has to be minimized or eliminated. In this respect, the decarboxylation process that removes oxygen from fatty acids in the form of CO2 appears to be more advantageous than hydrodeoxygenation that removes oxygen in the form of H2O.3−6 Almost all the decarboxylation studies were done using fatty acids without solvent or dissolved in hydrocarbon solvent, and the best results were achieved using noble metals (e.g., Pt/C, Pd/C) as catalysts.7−11 Using sub- and supercritical water for processing biomassderived products is considered an efficient and environmentally benign method that capitalizes on the extraordinary properties of water as a reactant and reaction medium and provides the ideal conditions for ionic/free-radical reactions. In subcritical waterbased processes, water is kept in a liquid state by applying pressure greater than the pressure of vapor at the reaction temperature. In this way, the energy consumed for the water to steam phase change (the latent heat of water vaporization is © XXXX American Chemical Society

2.26 MJ/kg) is avoided, which significantly reduces the energy requirements for hydrothermal processes.12 There is a growing potential for using hot compressed water as a reaction medium for catalytic decarboxylation of fatty acids.13−18 Subcritical water was proven to be a favorable environment for hydrolysis of triglycerides without using alkali or acid catalysts.19−22 Decarboxylation of the produced fatty acids into hydrocarbons can ideally be accomplished in the same aqueous medium without prior separation. Watanabe et al. demonstrated that stearic acid can be converted into alkanes in supercritical water at 400 °C for 30 min using NaOH and KOH as catalysts.13 However, the selectivity to hydrocarbons reported for this process was low (90%) toward the respective decarboxylation products for saturated fatty acids. Savage et al. also demonstrated that activated carbon alone possesses catalytic activity for hydrothermal decarboxylation of palmitic and oleic acids.16 The experiments for the above studies were conducted in batch reactors for 3 h without H2 added, and deoxygenation products from oleic acid contained both alkanes and alkenes. This indicates that an in situ formation of H2 occurs during the deoxygenation process, most likely from gasification of oleic acid itself, yielding H2 and aromatic/cyclic compounds. The amount of H2 produced is insufficient for complete hydrogenation of oleic acid and the resulting hydrocarbons. Received: February 7, 2015 Revised: April 8, 2015

A

DOI: 10.1021/acs.energyfuels.5b00308 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Figure 1. Experimental setup for rapid hydrothermal deoxygenation of oleic acid. the overall length of 23 in. and inside diameter of 9/16 in. equipped with a piston separator was used as a feeder. Formic acid (1% v/v) was preheated in the furnace, mixed with oleic acid, and delivered into the reactor with Pump 2. Reaction products were cooled down in a water heat exchanger. The liquid products were collected for mass balance calculations and further analyses. The volume flow rate of the gaseous products was measured periodically with a volumetric buret filled with water and a stopwatch. The samples of the gaseous products were taken from the buret with a syringe and analyzed with a gas chromatograph. The pressure inside the reactor was regulated with the back pressure regulator and kept close to 3500 ± 10 psi (24.1 MPa) for all the experiments. The ratio of oleic acid to 1% v/v formic acid was kept at 0.2:1 mL/min for all the experiments, i.e., oleic acid entered the system at 6.34 × 10−4 mol/min. The temperature inside the reactor was measured with two thermocouples equipped with controllers; one thermocouple was installed in the middle of the reactor, the other, immediately after the reactor. The experiments were conducted at six different temperatures: 350 ± 2 °C, 360 ± 2 °C, 370 ± 2 °C, 380 ± 2 °C, 390 ± 2 °C, and 400 ± 2 °C. The pressure in the system was measured with the pump controllers before the reactor and with PGS-35B-5000 general service pressure gauge (Omega, USA) before the back pressure regulator. The residence time in the reactor was adjusted to 21 min. Weight hourly space velocity (WHSV) was 4 h−1 for all the experiments and was calculated using eq 1:

The use of batch reactors takes a considerable amount of time to complete the conversions and makes it problematic to scale up the process. We attempted to develop and investigate a rapid and inexpensive continuous flow process that could be readily scaled up and used for renewable fuel production via hydrothermal deoxygenation of biomass with high lipid content. We used activated carbon as a catalyst in our process because it could be an attractive low-cost option for this type of conversion provided that it is sufficiently active under the conditions studied. The novelties of the proposed process are (i) using a hydrothermal medium for deoxygenation of fatty acids; (ii) using a continuous flow process that significantly reduces the reaction time and makes the process scalable; (iii) using activated carbon as a low-cost decarboxylation catalyst; (iv) using formic acid as an in situ source of H2 for hydrogenation of the intermediate products. To the best of our knowledge, there has been no report on hydrothermal deoxygenation of fatty acids using activated carbon as a catalyst and formic acid as an in situ source of H2 in a continuous flow process. In this article, we report on the rapid hydrothermal catalytic deoxygenation of oleic acid in a continuous flow reactor loaded with low-cost granulated activated charcoal.

WHSV =

2. EXPERIMENTAL SECTION

feed mass flow, g/h catalyst mass, g

(1)

The experiments started with pumping 1% v/v formic acid into the reactor until the desired temperature and pressure were reached and stabilized. After that, the other pump started delivering oleic acid into the reactor through the piston feeder. A typical experiment was conducted for 6 h, 40 min until the 81 mL piston feeder was empty. The liquid reaction products and water were collected in a 500 mL flask. The obtained products were separated from water, quantified, and analyzed with a gas chromatograph−mass spectrometer (GC/MS). After each experiment, the system was flushed with pure deionized water at 10 mL/min for at least 1 h. Two series of control experiments were conducted before the main experimental part. The first series was carried out with deionized water only to determine whether the gas formation other than from formic or oleic acids occurs during the deoxygenation process at different reaction temperatures (350−400 °C). The second series was carried out with oleic acid and water but without formic acid added to water to determine whether hydrogenation of oleic acid occurs without a donor of H2 at different reaction temperatures (350−400 °C). 2.2. Product Separation and Analyses. 2.2.1. Analyses of the Catalyst. Brunauer−Emmett−Teller (BET) surface area and pore

2.1. Materials and Methods. Oleic acid lab grade >97%, formic acid reagent grade 88%, and granulated activated charcoal Darco were purchased from Fischer Scientific USA. All the chemicals were used as received. The oleic acid was loaded into the piston feeder and used as a feedstock in the study. Formic acid (1% v/v) was added to deionized water and served as an additional source of H2 during the hydrothermal hydrogenation process. The granulated activated charcoal was sieved to remove particle sizes